Holistic Approach for Dimensioning and Optimization of Extrusion Dies
Background of this thesis is the need to extend the state of the art for the dimensioning of extrusion dies by automated approaches. For the extrusion and production process of semi-finished products, the die geometry has a significant influence on the polymer processing. In this context, the die is used in an extrusion process to form a semi-finished product. Therefore, an extruder supplies the die with molten thermoplastic material. The resulting product quality is significantly influenced by the flow phenomena in the extrusion die. Here, the dimensioning of the flow channel geometry is dependent on many requirements and significantly on the rheological properties of the polymer melt. Linked to these facts, a holistic approach for the automated optimization of extrusion dies has been developed. The main idea is the coupling of an one-dimensional analytical calculation approach for the pre-dimensioning of die geometries with a three-dimensional iteratively working numerical simulation. The general approach is formulated universally and is specified and tested exemplarily for spiral mandrel dies extruding for example plastic pipes, tubes and blown films.The created holistic approach takes into account all relevant steps of the dimensioning process and considers different parts of the die assembly. The fundament is the consequent usage of three-dimensional parametric geometry models. These models include geometric restrictions and manufacturing knowledge. For the pre-dimensioning, one-dimensional analytical calculation models are used. After pre-defining the state variables and the material properties, the geometry of the basic flow channels is calculated. The calculation of an initial geometry of a spiral mandrel die is based on solving a network of coupled one-dimensional analytical equations. Therefore, state variables and basic geometrical dimensions have to be defined to directly calculate an initial spiral depth and gap width. In addition, a percentage bandwidth is derived within a coupled optimization algorithm is able to change geometry parameters (spiral depth and gap width).Based on the geometry parameters and the bandwidth calculated using the pre-dimensioning, an automatic optimization takes place. Therefore, three-dimensional numerical flow simulations are used. Basis for the discretization is a three-dimensional parametric CAD model of the die including geometrical and manufacturing restrictions. A multiple criteria genetic optimization algorithm is coupled with a three-dimensional numerical flow simulation to optimize the die geometry by using defined quality and exclusion criteria calculating new parameter sets to change the geometry. Based on the analysis of general requirements of extrusion dies, these quality and exclusion criteria are defined and validated by comparative simulations.A manual detail optimization and virtual testing of the resulting die geometry follow this process. In this way, the whole extrusion die assembly is considered and interactions between different die regions are taken into account. Therefore, the possibility is given to change local areas of the flow channels. Exemplary the approach is used to optimize geometry regions with low wall shear stresses and flow velocities in the feed region of a spiral mandrel die. By analyzing the flow history along specific path lines, regions in the feeding section of the die are identified and associated geometry elements are changed. The optimization method is validated by practical experiments extruding blown film with an optimized and a reference spiral mandrel die. By studying product-changing processes, using spectral photometry, the amount of purging material is measured. The result is a significantly reduced amount of purging material for the optimized die geometry – and linked to that a significant reduction of residence time differences at the outlet of the die – compared to the reference die. In addition, a possibility is given to analyze different process conditions based on the simulation model of the complete die assembly by the usage of the virtual testing approach. The influence of material properties, mass flow rate and dissipative heating has been studied. Therefore, the process window within the die assembly could be used (range of materials with different rheological properties, range of mass flow rates) is defined. It has been found out that the heat transfer between polymer melt and the steel parts of the die assembly has a significant influence on the temperature distribution.
An essential decrease of calculation times and iteration numbers of the developed automated optimization approach is proven comparing to other theses. Parallel to that, an increase of the geometrical degrees of freedom is reached. The numerical simulations are validated by analyzing the results of practical experiments using manufactured die prototypes extruding blown film. Furthermore, the defined quality criteria are confirmed by temperature dependent simulations. It was figured out that the used criteria for the automated isothermal based optimization process also lead to an equal dissipative warming of the polymer melt during distribution in the die. Finally, the capability to implement a heat discharge for increasing mass flow rates was shown.Lesen Sie die deutsche Zusammenfassung auf Kunststoffe.de
Extrusion, Spiral-Mandrel-Dies, Design of Extrusion Dies, Computer Aided Optimization, Extrusion Die Optimization, Design of Spiral-Mandrel-Dies, Extrusion-Dies, holistic Optimization Algorithm, genetic Optimization, Virtual Testing
Institute / chair: Fakultät für Ingenieurwissenschaften, Abteilung Maschinenbau und Verfahrenstechnik der Universität Duisburg-Essen
Technical consultant for expert services: Prof. Dr.-Ing. Johannes Wortberg (Betreuer), Prof. Dr.-Ing. Volker Schöppner
Publication year: 2015
Provider: Wissenschaftlicher Arbeitskreis Kunststofftechnik (WAK) / Kunststoffe.de
About the PhD Dissertation Data Base
In cooperation with the Wissenschaftlicher Arbeitskreis Kunststofftechnik (WAK), we provide free access to current dissertations in the field of plastics technology.