1. Αποθετήριο Παραδοτέων Υποτρόφων Δράσεων ΕΣΠΑ Τριτοβάθμιας Εκπαίδευσης 2017-2023
2. Όλο το Αποθετήριο
3. Ερευνητικές ομάδες

Τύπος:	Κεφάλαιο σε πρακτικά συνεδρίου
Τίτλος:	Shape optimization of 3D diaphragm pumps using the continuous adjoint approach to the cut-cell method
Συγγραφέας:	[EL] Σαμούχος, Κωνσταντίνος[EN] Samouchos, Konstantinos [EL] Καψούλης, Δημήτριος[EN] Kapsoulis, Dimitrios [EL] Τρομπούκης, Ξενοφών[EN] Trompoukis, Xenofon [EL] Γιαννάκογλου, Κυριάκος[EN] Giannakoglou, Kyriakos
Ημερομηνία:	09/07/2019
Περίληψη:	This paper is concerned with tools that can be used for the design/optimization of 3D valveless and valved diaphragm pumps. These pumps comprise a main chamber, an inlet and outlet duct and a periodically moving diaphragm that causes the unsteady/periodic fluid flow within the device; the inlet and outlet ducts might either be valveless diffusers or tubes of constant cross-section with valves (valved pumps). Obtaining the desired net mass-flow rate per period with the minimum (hopefully zero) back-flow at the exit (especially, in the absence of valves) are the usual objectives. The minimization of viscous losses within some pump’s components such as inlet or outlet valved ducts could be an extra target. Regarding CFD analysis, apart from the cut-cell method mentioned in the title, an in-house flow solver using body-fitted unstructured grids is also used. Both codes are based on the finite volume approach by sharing the same numerical features and run on GPU-clusters (the former) and on many-processor platforms (the latter). To support gradient-based optimization, with either CFD tool, the continuous adjoint method computing the gradient of the objectives with respect to the design variables has been developed and programmed. According to the proposed parameterization scheme, all design variables are related to the diaphragm geometry and its periodic motion. An alternative optimization method that was also made available for the needs of this paper is a (gradient-free) evolutionary algorithm, assisted by metamodels (on-line trained artificial neural networks) and the principal component analysis for efficiently searching the design space. Using this tool, many-objective optimizations can be performed to compute the front of non-dominated solutions. This paper focuses on particular features of the analysis and optimization tools for this type of diaphragm pumps, including micropumps for medical applications.
Γλώσσα:	Αγγλικά
Τόπος δημοσίευσης:	Singapore
Σελίδες:	14
ISSN:	2374-3948
Θεματική κατηγορία:	[EL] Επιστήμη Μηχανολόγου Μηχανικού[EN] Mechanical Engineering
Λέξεις-κλειδιά:	Cut-cell method; Adjoint method; Diaphragm pumps; Evolutionary Algorithm
Κάτοχος πνευματικών δικαιωμάτων:	© by the author(s)
Διατίθεται ανοιχτά στην τοποθεσία:	http://velos0.ltt.mech.ntua.gr/research/confs/3_151.pdf https://www.sci-en-tech.com/ICCM2019/ICCM2019-Proceedings.pdf https://www.sci-en-tech.com/ICCM2019/ICCM2019-Presentations.pdf http://velos0.ltt.mech.ntua.gr/research/pubs.html#link_conferences
Ηλεκτρονική διεύθυνση του τεκμηρίου στον εκδότη:	http://velos0.ltt.mech.ntua.gr/research/confs/3_151.pdf https://www.sci-en-tech.com/ICCM2019/ICCM2019-Proceedings.pdf
Ηλεκτρονική διεύθυνση περιοδικού:	https://sci-en-tech.com/ICCM2020/Paper%20Proceedings%20for%20Download.pdf
Τίτλος πηγής δημοσίευσης:	10th International Conference on Computational Methods Full Paper Proceedings
Τόμος:	6
Σελίδες τεκμηρίου (στην πηγή):	171-184
Όνομα εκδήλωσης:	10th International Conference on Computational Methods ICCM
Τοποθεσία εκδήλωσης:	Singapore
Ημ/νία έναρξης εκδήλωσης:	09/07/2019
Ημ/νία λήξης εκδήλωσης:	13/07/2019
Σημειώσεις:	Research presented in this paper has been funded by the Business Plan “Development of Human Resources, Education and Lifelong Learning” entitled “Support Researchers with Emphasis on Young Researchers” with the co-financing of Greece and the European Union. The project title is “Design-Optimization of Diaphragm Pumps under Operational/Manufacturing Uncertainties using the Cut-Cell Method and Polynomial Chaos Expansion”.
Εμφανίζεται στις συλλογές:	Ερευνητικές ομάδες