The present work is concerned with the automated optimization of the design of PVC profile extrusion dies. The aspects of the research are the parameterization techniques, the development of suitable optimization strategies, and the determination of material parameters from experimental studies. Due to the aim of this work, namely to apply automatic optimization techniques to industrially relevant PVC profile dies, the necessary tools have to be developed such, that these are capable of dealing with profile dies with a comparatively complex shape while being still efficient. The starting point of this work is a collection of CFD tools based on an FE flow analysis performed on so-called 2d die-slices, which represent 2d cross-sectional cycle is set-up, in which shape optimization is applied to each die-slice geometry. As objective function the flow rate fraction distribution is used leading to a flow balancing. Parameterization tools are developed that deal with the required automatic geometry modification within the optimization cycle. For this purpose a number of design variable (DV) types are pre-defined, which enable to prescribe DVs on the die-slice geometry before entering the optimization cycle. Due to the comparatively high number of DVs and the intended industrial application, an efficient optimization strategy is necessary. Several methods are looked at and a suitable strategy is developed. This strategy performs well regarding the required manual time to set-up and the required computational time. A comparison with a manual die design process confirms that the resulting geometry is acceptable.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:636911 |
| Date | January 2005 |
| Creators | Ettinger, H. J. |
| Publisher | Swansea University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0116 seconds