ISBN 9783843936545

978-3-8439-3654-5, Reihe Ingenieurwissenschaften

Roland Siegbert
Quantification of Viscoelastic Die Swell as a Design Objective in the Numerical Design of Profile Extrusion Dies

166 Seiten, Dissertation Rheinisch-Westfälische Technische Hochschule Aachen (2018), Softcover, A5

Zusammenfassung / Abstract

Individualized production, which is a major goal of many high-wage countries, describes a production process in which all elements of a production system are designed in such a way that they enable a high level of product variety at mass production costs. This thesis demonstrates recent advances in the individualized production with die-based manufacturing processes, namely plastics profile extrusion.

The chosen approach is aiming at the use of numerical die design methods, where the design procedure itself relies on nonlinear numerical optimization. and a spline-based geometry library. All components interact automatically requiring only little user interaction; thus, a completely independent optimization cycle is achieved. Numerical optimization helps to reduce — or even eliminate — the so far very time-consuming manual reworking steps of an initial die or process design. These manual reworking steps are a major cost factor when it comes to individual production. Their abolishment through the presented numerical approaches therefore is an important step towards the concept of individualized production, which is a major goal of the thesis being embedded in the Cluster of Excellence ”Integrative Production Technology for High-Wage Countries”. In the context of of this thesis it was possible to identify several typical design criteria mirroring the approach of experienced die designers.

The developed optimization framework XEOS establishes a modular, robust, efficient and extendable optimization process. By taking successful ideas from a first generation framework, which has been developed at the Chair for Computational Analysis of Technical Systems, into account the simulation of free surfaces is now feasible and allows for the implementation of more sophisticated objective functions. Further, a new spline fitting method has been developed and fine-tuned to the requirements imposed by extrudate die swell behavior. With little impact on the overall computational time, the new methods enable efficient simulation and die swell evaluation in academic and an industrial use cases, which can be utilized within the new optimization framework.