Bekannte Bauingenieure
Dresdner Bauseminar
Ehren- und Festkolloquien
Ehrungen und Preise
... weitere
Weblog abonnieren
beyer bau
Das Weblog der Dresdner Bauingenieure

Über "The use of approximation concepts in the structural synthesis of stochastic dynamical systems" spricht am 15. November 2006 um 10.30 Uhr Prof. Ph.D. Héctor Jensen vom Department of Civil Engineering der Federico Santa Maria University Valparaiso, Chile im
Sitzungszimmer Raum 67 des Beyerbaus.
Der Vortrag findet (der Titel lässt es erahnen) in englischer Sprache statt, die nachfolgende Zusammenfassung ist somit eine gute Vorbereitung auf das SFB-528-Seminar:

"This seminar concentrates on the structural optimization of linear and non-linear systems under stochastic excitation. Formulations based on response statistics (RSBO) and reliability measures (RBO) are discussed. For RSBO formulations the stochastic response is characterized by the first two statistical moments, which are computed by statistical linearization techniques in the case of non-linear systems. For an efficient solution of the optimal design process the original structural optimization problem is replaced by a sequence of linearized explicit sub-optimization problems. The sub-optimization problems are constructed by using a convex first-order approximation of the objective and constraint functions. In the case of RBO the reliability of the system is defined in terms of first excursion probabilities. Failure probabilities are approximated locally during the optimization process, and they are represented using approximate functions dependent on the design variables. The local approximations are then used to define a series of local explicit optimization problems, which are solved in an efficient manner. The applicability and feasibility of the proposed design process is demonstrated in several examples where the methodology is applied to linear systems and systems with nonlinearity of hardening and hysteretic type. The numerical applications include MDOF-systems and FEM-systems under stochastic loading. The use of approximation concepts dramatically reduces the number of exact dynamic and stochastic analyses as well as reliability estimations required during the optimization process. The effects of the nonlinearity on the general performance of the final designs as well as some engineering implications of the results obtained are discussed.

resident of
powered by Antville powered by Helma
AGBs xml version of this page