Deep learning study of induced stochastic pattern formation in the gravure printing fluid splitting process / Pauline Brumm in JOURNAL OF COATINGS TECHNOLOGY AND RESEARCH, Vol. 20, N° 1 (01/2023)  

[article]  inJOURNAL OF COATINGS TECHNOLOGY AND RESEARCH > Vol. 20, N° 1 (01/2023) . - p. 51-72 Titre : Deep learning study of induced stochastic pattern formation in the gravure printing fluid splitting process Type de document : texte imprimé Auteurs : Pauline Brumm, Auteur ; Nicola Ciotta, Auteur ; Hans Martin Sauer, Auteur ; Andreas Blaeser, Auteur ; Edgar Dörsam, Auteur Année de publication : 2023 Article en page(s) : p. 51-72 Note générale : Bibliogr. Langues : Américain (ame) Catégories : Algorithmes Apprentissage Héliogravure Hydrodynamique Intelligence artificielle Reconnaissance de formes Index. décimale : 667.9 Revêtements et enduits Résumé : We use deep learning (DL) algorithms for the phenomenological classification of Saffman-Taylor-instability-driven spontaneous pattern formation at the liquid meniscus in the fluid splitting in a gravure printing press. The DL algorithms are applied to high-speed video recordings of the fluid splitting process between the rotating gravure cylinder and the co-moving planar target substrate. Depending on rotation velocity or printing velocity and gravure raster of the engraved printing cylinder, a variety of transient liquid wetting patterns, e.g., a raster of separate drops, viscous fingers, or more complex, branched liquid bridges appear in the printing nip. We discuss how these patterns are classified with DL methods, and how this could serve the identification of different hydrodynamic flow regimes in the nip, e.g., point or lamella splitting. Note de contenu : - EXPERIMENTAL : Experimental setup with high-speed camera - Variety of observed patterns - DATA ANALYSIS : Fundamentals of DL - Computer vision workflow - RESULTS : Test accuracies and training duration - Confusion matrices - Class probabilities - CAMs for 3D-CNN architecture models - Table 1 : Confusion matrix for a binary classification problem - Table 2 : List of used Python version, package manager, and Python libraries - Table 3 : Test accuracies for all DL models after 20 training epochs - Table 4 : Test accuracies for different optimizations and numbers of frames after 20 training epochs for model #2 (3D-CNN, 224 px x 224 px, 3-class-model) - Table 5 : List of videos used for exemplary snapshots in this paper with corresponding printing parameters and label - Table 6 : Confusions of all trained DL models on the test data set DOI : https://doi.org/10.1007/s11998-022-00687-x En ligne : https://link.springer.com/content/pdf/10.1007/s11998-022-00687-x.pdf?pdf=button% [...] Format de la ressource électronique : Pdf Permalink : https://e-campus.itech.fr/pmb/opac_css/index.php?lvl=notice_display&id=38829 [article]

Réservation

Réserver ce document

Exemplaires (1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
23928	-	Périodique	Bibliothèque principale	Documentaires	Disponible

---

1 (1 - 1 / 1)