Development of a Process for Thermal and Mechanical Modelling of Screw-Driven Pellets Extrusion

The overall goal of the thesis project is to develop a process for thermal and mechanical modelling of the screw-driven pellets extrusion process, and applying the model results to design extruder temperature and flow rate controllers. The proposed extruder is designed for metal 3D printing. The device demonstrates great potential in tackling some of the major issues faced by the metal additive manufacturing community. It eliminates the use of metal powder for workplace and workers safety. It is able to produce end-use parts with industrial grade mechanical and microstructural properties. It utilizes low cost metal-loaded polymer pellets as feedstock. However, the application is only possible when the extruder has an accurate and responsive control system. Design of the extruder controller depends on a thorough understanding of the extrusion process. While a variety of polymer extrusion models exist in literature, most of them approximate the feedstock as a Newtonian fluid and make simplified assumptions about the pressure and temperature profile of the feedstock. The accuracy of the results are not sufficient for 3D printer control. Even less literature exists studying the extrusion process of metal injection molding machines, as an accurate flow control is not necessary for injection molding processes. To fill the gap in literature, the objectives of the thesis involve developing a heat transfer and a flow rate model that realistically characterize the screw-based extrusion process, and applying the models to design a comprehensive extruder temperature and flow rate control system. The models are validated with the existing extruder prototype and PLA feedstock pellets. While the model details might be different for different materials and extruder geometries, the modelling process should be universally applicable to all kinds of feedstock, including metal-loaded polymer pellets. A heat transfer model is proposed for the extruder prototype using a finite volume method. The goal of the model is to simulate the extruder and the feedstock temperature distribution given the heating and cooling system input. The model divides the extruder and the feedstock into 36 different control volumes. Conservation of energy and multi-node heat transfer equations are used to simulate the heat transfer between each control volume. The model is able to predict the extruder and the feedstock temperatures within 5°C compared to the experiment data. The model can be used to optimize the heater and cooling water input to provide an ideal thermal processing condition for the feedstock. A steady state output mass flow rate model is developed based on a simplified polymer extrusion model from literature. It incorporates a shear rate dependent polymer viscosity model and a calibrated feedstock pressure profile to increase model accuracy. The feedstock temperature distribution simulated in the heat transfer model is used to calculate various temperature dependent material properties. The model yields a logarithmic-like relationship between output mass flow rate and screw rotation speed. It reduces the error in the original simplified model by more than 50%. A post flow model is developed upon the steady state flow rate model results. The process utilizes a polymer compressibility model and the calculated extruder operating pressures to predict the amount of leaked extrudate after the screw stops rotating. A controller is proposed to add screw retraction at the end of each extrusion to eliminate post flow. It reduces the amount of leaked extrudate by more than 90%, as shown in the experiment. Finally a dynamic output mass flow rate model is presented. A first-order approximation is used to model the dynamic response of output flow rate with respect to change of screw rotation speeds. Results from both the steady state flow rate model and experiments are used to determine the constants within the dynamic model. A proportional controller is proposed to dynamically control the output mass flow rate. Further experiments need to be performed to design and validate the controller. The thesis is successful in developing a process for modelling and controlling desktop screw extruder. The post flow model and the dynamic flow rate model provide valuable insights on how to accurately control the extruder output for 3D printing applications. In the future, the modelling process can be applied to feedstock materials, and serve as a general guidelines for future screw extruder design