Graduiertenkolleg 1462
Refine
Document Type
- Doctoral Thesis (3) (remove)
Keywords
- Bayes-Inferenz (1)
- Bewertung (1)
- Bewertungsmethode (1)
- Coupled models (1)
- Evaluation methods (1)
- Gekoppelte Modelle (1)
- Ingenieurbau (1)
- Kriechen (1)
- Model quality, Model error estimation, Kinematical model, Geometric non-linearity, Finite Element method (1)
- Modellqualität, Modellfehlerschätzer, Geometrisch nicht-lineare Berechnung, Kinematik Modell, Finite Elemente Methode (1)
Year of publication
- 2012 (3) (remove)
Methods for model quality assessment are aiming to find the most appropriate model with respect to accuracy and computational effort for a structural system under investigation. Model error estimation techniques can be applied for this purpose when kinematical models are investigated. They are counted among the class of white box models, which means that the model hierarchy and therewith the best model is known. This thesis gives an overview of discretisation error estimators. Deduced from these, methods for model error estimation are presented. Their general goal is to make a prediction of the inaccuracies that are introduced using the simpler model without knowing the solution of a more complex model. This information can be used to steer an adaptive process. Techniques for linear and non-linear problems as well as global and goal-oriented errors are introduced. The estimation of the error in local quantities is realised by solving a dual problem, which serves as a weight for the primal error. So far, such techniques have mainly been applied in
material modelling and for dimensional adaptivity. Within the scope of this thesis, available model error estimators are adapted for an application to kinematical models. Their applicability is tested regarding the question of whether a geometrical non-linear calculation is necessary or not. The analysis is limited to non-linear estimators due to the structure of the underlying differential equations. These methods often involve simplification, e.g linearisations. It is investigated to which extent such assumptions lead to meaningful results, when applied to kinematical models.
Bridge vibration due to traffic loading has been a subject of extensive research in the last decades. The focus of such research has been to develop solution algorithms and investigate responses or behaviors of interest. However, proving the quality and reliability of the model output in structural engineering has become a topic of increasing importance. Therefore, this study is an attempt to extend concepts of uncertainty and sensitivity analyses to assess the dynamic response of a coupled model in bridge engineering considering time-dependent vehicular loading. A setting for the sensitivity analysis is proposed, which enables performing the sensitivity analysis considering random stochastic processes. The classical and proposed sensitivity settings are used to identify the relevant input parameters and models that have the most influence on the variance of the dynamic response. The sensitivity analysis exercises the model itself and extracts results without the need for measurements or reference solutions; however, it does not offer a means of ranking the coupled models studied. Therefore, concepts of total uncertainty are employed to rank the coupled models studied according to their fitness in describing the dynamic problem.
The proposed procedures are applied in two examples to assess the output of coupled subsystems and coupled partial models in bridge engineering considering the passage of a heavy vehicle at various speeds.
Ziel dieser Arbeit ist die Entwicklung von Methoden, mit denen die Prognosequalität von Kriechmodellen des Betons bestimmt werden kann. Die Methoden werden in zwei Ausgangsszenarien unterschieden: die Bewertung ohne und die Bewertung mit Verwendung von spezifischen Versuchsdaten zum Kriechverhalten des Betons. Die Modellqualität wird anhand der Gesamtunsicherheit der prognostizierten Kriechnachgiebigkeit quantifiziert. Die Unsicherheit wird für die Kriechprognose ohne Versuchsdaten über eine Unsicherheitsanalyse unter Berücksichtigung korrelierter Eingangsparameter ermittelt. Bei der Verwendung experimenteller Daten werden die stochastischen Eigenschaften der Modellparameter mittels Bayesian Updating bestimmt. Die Bewertung erfolgt erneut basierend auf einer Unsicherheitsanalyse sowie alternativ mittels Modellselektion nach Bayes.
Weiterhin wird eine auf Graphentheorie und Sensitivitätsanalysen basierende Methode zur Bewertung von gekoppelten Partialmodellen entwickelt. Damit wird der Einfluss eines Partialmodells auf das Verhalten einer globalen Tragstruktur quantifiziert, Interaktionen von Partialmodellen festgestellt und ein Maß für die Qualität eines Gesamtmodells ermittelt.