University of West Attica
Σχεδιασμός για την προσθετική κατασκευή, πειραματική διερεύνηση, ευφυή προγνωστική μοντελοποίηση και βελτιστοποίηση της μηχανικής συμπεριφοράς εξαρτημάτων
Abstract
dc:descriptionAdditive Manufacturing (AM) has emerged as one of the most significant technological developments in modern manufacturing, offering unprecedented design freedom, the capability to fabricate highly complex geometries, and considerable reductions in product development time. Despite its rapid industrial adoption, the mechanical performance of additively manufactured components remains strongly dependent on numerous process parameters, highlighting the need for reliable optimization and predictive methodologies. The primary objective of this doctoral dissertation is to investigate the influence of the most critical additive manufacturing process parameters on the mechanical behavior of specimens subjected to three-point bending and to develop advanced predictive models capable of accurately estimating their mechanical properties. The research focuses on Fused Filament Fabrication (FFF/FDM) and Stereolithography (SLA) technologies using PLA, ABS, and photopolymer resin materials. The experimental investigation was based on the Taguchi Design of Experiments methodology employing orthogonal arrays, enabling the systematic evaluation of multiple process variables while significantly reducing the number of required experimental trials. Specimens were fabricated under various combinations of processing parameters and tested according to ASTM D790 three-point bending standard. Multiple repetitions were conducted for each experimental condition, accompanied by detailed dimensional measurements and statistical assessment of experimental repeatability. The acquired experimental data were analyzed using Signal-to-Noise (S/N) ratio analysis, Analysis of Variance (ANOVA), multiple linear regression modelling, and Adaptive Neuro-Fuzzy Inference System (ANFIS) modelling. Furthermore, optimization procedures for the ANFIS membership functions were implemented and evaluated to improve the predictive capability of the developed intelligent models. The experimental findings demonstrate that printing orientation, infill pattern, infill density, layer thickness, thermal processing conditions, and post-processing parameters constitute the dominant factors affecting maximum bending stress, fracture strain, and flexural modulus. The analysis further highlights the pronounced anisotropic behavior of additively manufactured components and confirms the decisive influence of interlayer bonding quality on their overall mechanical performance. The comparative evaluation of the applied methodologies indicates that the Taguchi method provides an efficient framework for process optimization and factor significance assessment, whereas regression models offer reliable analytical prediction equations within the investigated design space. Among all examined approaches, the ANFIS models achieved the highest prediction accuracy by successfully capturing the highly nonlinear relationships between manufacturing parameters and mechanical properties. The principal contribution of this dissertation is the development of an integrated methodology that combines experimental design, statistical analysis, mathematical modelling, and artificial intelligence techniques for the prediction and optimization of the mechanical performance of additively manufactured components. The proposed framework provides a valuable decision-support tool for industrial additive manufacturing applications and establishes a solid foundation for the future development of intelligent process optimization systems.
Degree
thesis:*- Grantor dc:publisher
- University of West Attica
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Sagias, Vasileios
- Σαγιάς, Βασίλειος
Subjects
dc:subject × 40- Προσθετική κατασκευή
- Σχεδιασμός για προσθετική κατασκευή
- Μηχανικές ιδιότητες
- Σχεδιασμός πειραμάτων
- Μέθοδος Taguchi
- Ανάλυση παλινδρόμησης
- Μηχανική μάθηση
- Προσαρμοστικό νευρο-ασαφές σύστημα συμπερασμού
- Προβλεπτική μοντελοποίηση
- Βελτιστοποίηση διεργασίας
- Πειραματικός χαρακτηρισμός μηχανικών Ιδιοτήτων
- Πειραματική διερεύνηση μηχανικών ιδιοτήτων
- Additive manufacturing
- Design for Additive Manufacturing
- Mechanical properties
- Design of experiments (DOE)
- Taguchi method
- Regression analysis
- Machine learning
- Adaptive Neuro-Fuzzy Inference System
- ANFIS
- Predictive modeling
- Process optimization
- Experimental characterization of mechanical properties
- Experimental investigation of mechanical properties
- DfAM
- AM
- Επιστήμη Μηχανολόγου Μηχανικού
- Επιστήμες Μηχανικού και Τεχνολογία
- Μηχανολογία
- Μηχανολογική μηχανική
- Μηχανική Υλικών
- Επιστήμη υλικών, χαρακτηρισμός και δοκιμές
- Πολυμερή και Πλαστικά
- Mechanical Engineering
- Engineering and Technology
- Mechanics
- Materials Engineering
- Materials Science, Characterization and testing
- Polymers and Plastics
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/62771