Comparative study of reduced-order methods for geometrically nonlinear structures

Khurshudyan, Amalya

Reference : Comparative study of reduced-order methods for geometrically nonlinear structures


	Document type :	Dissertations and theses : Doctoral thesis
	Discipline(s) :	Engineering, computing & technology : Civil engineering
	Focus Areas :	Computational Sciences
	To cite this reference:	http://hdl.handle.net/10993/32867

Title :	Comparative study of reduced-order methods for geometrically nonlinear structures
Language :	English
Author, co-author :	Khurshudyan, Amalya [University of Luxembourg > Faculty of Science, Technology and Communication (FSTC) > Engineering Research Unit >]
Publication date :	11-May-2017
Institution :	University of Luxembourg, Dahlem, Luxembourg
Name of the degree :	Docteur De L’Universite Du Luxembourg En Sciences De l’Ingenieur
Number of pages :	111
Supervisor :	Zilian, Andreas
President of the jury :	Bordas, Stéphane
Member of the jury :	Baroli, Davide
	Legay, Antoine
	Belouettar, Salim
Keywords :	[en] Model Order Reduction ; Geometrical Non-Linearity ; Structural Analysis
Abstract :	[en] The dissertation is devoted to the comparison and development of techniques for model order reduction (MOR) of geometrically nonlinear elastic structures in the static limit. The MOR procedure works in the following way: the structure is first discretized into finite elements and a discretized system of algebraic equations is obtained, in which the stiffness matrix depends on the unknown vector. The system is then projected to a lower order space. The choice of the basis of the projection space is made according to the methods developed in the thesis. To this end, three techniques are developed here based on different choices of the basis functions. Comparative analysis of the suggested methods is carried out in the case of two-dimensional structures (Euler-Bernoulli beam, multi-span beam and frame). In order to be able to compare the results with those obtained by the MOR techniques which are developed, the benchmark problems which are examined are first solved analytically. Results of computations carried out in Python and are then discussed.
Permalink :	http://hdl.handle.net/10993/32867

File(s) associated to this reference

Fulltext file(s):

	File	Commentary	Version	Size	Access
Open access	AKhurshudyan_thesis.pdf		Author postprint	2.21 MB	View/Open

All documents in ORBi^lu are protected by a user license.

University of Luxembourg Library | Feedback | Legal notices