Datei im Netzwerk der BAM verfügbar ("Closed Access")
Filtern
Dokumenttyp
Sprache
- Englisch (2)
Referierte Publikation
- ja (2)
Schlagworte
- Artificial neural network (2) (entfernen)
Organisationseinheit der BAM
Directed energy deposition (DED) has been in industrial use as a coating process for many years. Modern applications include the repair of existing components and additive manufacturing. The main advantages of DED are high deposition rates and low energy input. However, the process is influenced by a variety of parameters affecting the component quality. Artificial neural networks (ANNs) offer the possibility of mapping complex processes such as DED. They can serve as a tool for predicting optimal process parameters and quality characteristics. Previous research only refers to weld beads: a transferability to additively manufactured three-dimensional components has not been investigated. In the context of this work, an ANN is generated based on 86 weld beads. Quality categories (poor, medium, and good) are chosen as target variables to combine several quality features. The applicability of this categorization compared to conventional characteristics is discussed in detail. The ANN predicts the quality category of weld beads with an average accuracy of 81.5%. Two randomly generated parameter sets predicted as “good” by the network are then used to build tracks, coatings,walls, and cubes. It is shown that ANN trained with weld beads are suitable for complex parameter predictions in a limited way.
Additive manufacturing, and therefore directed energy deposition, is
gaining more and more interest from industrial users. However, quality assurance for the components produced is still a challenge. Machine learning, especially using artificial neuronal networks, is a potential method for ensuring a high-quality standard. Based on process Parameters and monitoring data, part quality can be predicted. A further advantage is the ability to constantly learn and adopt to slight process changes.
First tests using artificial neural networks focus on the prediction of track geometry. The results show that even a small data set is enough to provide high accuracy in the predictions. In this work, an artificial neural network for the predictive analysis of relative density in laser powder cladding has been developed. A central composite experimental design is used to generate 19 data sets. Input variables are laser power, feed rate and powder mass flow. Cubes are built up where density is considered as a target value. Several neural networks are trained and evaluated with these data sets. Different topologies and initial weights are considered. The best network reaches a confidence level of around 90 % for the prediction of relative density based on the process parameters. Finally, the optimization of the generalization performance is investigated. To this purpose, methods of variation in error limit as well as cross-validation are applied. In this way, density is predictable by an artificial neural network with an accuracy of about 95 %.