University of Thessaly (UTH)
Design optimization of structures subjected to stochastic excitations
Abstract
dc:descriptionThis thesis focuses on the optimal design of structures under the influence of stochastic dynamic loads, with applications to the sizing and topology optimization of truss structures. Furthermore, the study investigates the fatigue lifetime prediction of structures under stochastic dynamic loading. Finally, the design optimization framework is extended to incorporate performance functions related to fatigue lifetime of structures subject to stochastic dynamic loading. Specifically, an innovative methodology for optimizing structures under stochastic dynamic excitations is proposed. The objective function related to the performance of the structure is constructed as a weighted sum of the variance of the structural response quantities of interest. Such response quantities are selected to be the displacements at the nodes of the system, stresses or strains developed in the structural parts, drift ratios etc. The excitations are modeled by white noise or filtered white noise processes. The variance of the response quantities are very efficiently estimated by solving the Lyapunov equation for the system. The adjoint method is developed to analytically estimate the sensitivities of the objective function with respect to the design parameters. This substantially increases the computational efficiency of the proposed methodology. The required computational effort for estimating the derivatives of the objective function is shown to be independent of the number of the design variables. Furthermore, the proposed methodology is extended to handle in the optimization the formulation of the response in the modal space. This exploits the benefits of using limited number of modes for the estimation of the system response The proposed design optimization of dynamic systems is applied for addressing the problem of size and topology optimization of truss structures. A specific class of two dimensional truss structure subject to Gaussian white noise excitation is considered to demonstrate the theoretical developments. Using different types of performance functions, such as the weighted sum of the nodal displacements and the weighted sum of the stresses developed in the truss members, the structures is optimized in size and topology in order to find the optimal configuration that optimizes the performance. The results are compared to available design optimization results of the truss structures subject to deterministic static loading. The analysis in modal space is also applied on the truss structure, in order to explore the effect of using limited number of contributing modes in the estimation of the response and finally investigate their effect on optimal design. A novel methodology for the fatigue reliability assessment of randomly vibrating multi-degree of freedom systems is also proposed within the coupled response-degradation model. The fatigue process in the system components is quantified by the fatigue crack growth equations which - via the stress range - are coupled with the system response. Simultaneously, the system dynamics are affected by the fatigue process through its stiffness degradation, so that it provides the actual stress values to the fatigue crack growth equation. In addition to the general coupled response-degradation analysis, its special case of non-coupled fatigue crack growth is treated as well, for the wide-band stationary applied stress by the use of its first four spectral moments and the approximate, empirically motivated, Dirlik’s probability distribution for the stress range. The proposed methodology is applied on multi degree of freedom spring mass chain like structure and the results for the different approaches are compared. Both, the general analysis and the illustrating examples elaborated provide the route to the fatigue reliability estimation in complex-hierarchical vibratory systems under random loading. Finally, the proposed methodologies for optimal design and fatigue lifetime prediction for structure subjected to stochastic dynamic excitations are combined, in order to form an innovative framework for design optimization of structures based on fatigue lifetime related performance function. In particular, the performance function for the optimization problem is defined as the weighted sum of the fatigue lifetimes of the structural parts.
Degree
thesis:*- Grantor dc:publisher
- University of Thessaly (UTH)
- Year dc:date
- 2010
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Perros, Kyriakos
- Πέρρος, Κυριάκος
Subjects
dc:subject × 22- Βέλτιστος σχεδιασμός
- Στοχαστική Διέγερση
- Στοχαστική Ταλάντωση
- Δικτύωμα
- Ροπή δεύτερης τάξης
- Κόπωση
- Υποβάθμιση ακαμψίας
- Ανάπτυξη ρωγμής
- Πιθανότητα αστοχίας
- Design optimization
- Stochastic excitation
- Random vibration
- Truss structure
- Second οrder moment
- Fatigue
- Stiffness degradation
- Crack growth
- Failure probability
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Μηχανολόγου Μηχανικού
- Engineering and Technology
- Mechanical Engineering
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/19724