National Technical University of Athens (NTUA)
Σχεδιασμός για κατασκευή μηχανολογικών τεμαχίων με γλυπτές επιφάνειες με βάση μορφολογικά χαρακτηριστικά
Abstract
dc:descriptionThis thesis studies the optimization of design and manufacture (machining) of parts consisting of sculptured surfaces their form being a positive or negative volume as typically used in the plastics forging and casting industry. First, extending the concept of morphological characteristics (features) is proposed by analogy to detailed analytical classical surfaces to sculptured surfaces In order for this transition to be meaningful the solid nature of the morphological characteristics is retained and it should not change to the surface nature. Morphological features are defined based on their functionality in the product. Each of these characteristics is defined based on parametric wireframe models that include constraints. According to this approach existing CAD technology is being exploited which adopts the concepts of parameterization and constraints and is fully implemented for curves but not for surfaces or solids. Each morphological feature is ultimately created as a solid with the standard Boolean and other solid modeling operators (extrude loft revolve etc.) in a CAD system. The result is a product designed so that its form can be easily changed by changing the values of the parameters. Secondly, the thesis proposes creation of a decision making system concerning CNC machining issues for parts with sculptured surfaces. These decisions refer to roughing and finishing tool diameter selection and the necessity or not of the use of a semi finishing process. The factors associated with these decisions are the volume of material to be removed the complexity of the surface to be created the maximum curvature and the relative importance of machining time in comparison to surface quality. Moreover for each of the three machining phases (roughing semi finishing finishing) several parameters that define machining strategy are decisive e.g. vertical stepdown value horizontal overlap value etc and their values enable the given time-quality balance to be obtained or to be approximated as close as possible. Given the inherent approximate nature of these decisions the decision system was implemented as a fuzzy system and leads the user to the choice of machining strategy. Thirdly, the thesis proposes the creation of neural meta models linking the parameters of morphological features with functional characteristics and characteristics of its manufacturing process as well. Neural networks were trained with a number of workpiece variations that were created by changing its morphological characteristic values. In generating training examples the decision making tool was applied concerning machining strategy and the computational capabilities of the CAD modeling environment were employed as well. Thus, neural networks replace both the decision making tool and the CAD environment which is why they are characterized as meta models. Next the calculated parameters were used in the objective function of a standard genetic algorithm which aimed at finding the best values of the model morphological characteristics based on functionality and manufacturing criteria. The relative importance of the criteria is defined by the user using weights. Due to this method being as accurate as the neural meta models special software to define the optimal architecture was used in order to obtain acceptably accurate meta models. The results of the thesis are directly applicable to the design and construction of mechanical parts with sculptured surfaces and especially molds.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2009
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Λυπαρή, Βιργινία
Subjects
dc:subject × 16- Γλυπτές επιφάνειες
- Σχεδιασμός με βάση τα μορφολογικά χαρακτηριστικά
- Παραμετροποίηση
- Ασαφές μοντέλο
- Επιλογή εργαλείου
- Στρατηγική φρεζαρίσματος
- Freeform surface
- Feature-based design
- Parameterization
- Fuzzy models
- Tool selection
- Milling strategy
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Μηχανολόγου Μηχανικού
- Engineering and Technology
- Mechanical Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/26916