Thomas Vogel Book order

Die Entwicklung selbst-adaptiver Software erfordert eine „Adaptation Engine“, die durch Feedbackschleifen die Software steuert und anpasst. Diese Anpassung wird oft durch Laufzeitmodelle beschrieben, die die Software repräsentieren, sowie durch Aktivitäten wie Analyse und Planung. Um das Zusammenspiel zwischen Laufzeitmodellen und Aktivitäten systematisch zu erfassen, wurden Megamodelle zur Laufzeit vorgeschlagen. Ein Megamodell zur Laufzeit ist ein spezielles Laufzeitmodell, dessen Elemente Aktivitäten und andere Laufzeitmodelle sind. Es erfasst das Zusammenspiel zwischen diesen Elementen und die Aktivierung sowie Ausführung der Aktivitäten. In diesem Artikel wird eine Modellierungssprache für ausführbare Megamodelle zur Laufzeit vorgestellt, die als EUREMA bezeichnet wird. Diese Sprache erleichtert die Entwicklung selbst-adaptiver Software durch einen modellgetriebenen Ansatz, der eine domänen-spezifische Modellierungssprache und einen Laufzeit-Interpreter für Adaptation Engines umfasst. EUREMA Megamodelle werden nicht nur in der Spezifikationsphase, sondern auch zur Laufzeit genutzt, um Feedbackschleifen direkt auszuführen und dynamisch anzupassen. Dadurch unterstützt EUREMA die Entwicklung selbst-adaptiver Software durch die explizite Spezifikation von Feedbackschleifen, Laufzeitmodellen und Adaptionsaktivitäten auf einer höheren Abstraktionsebene. Zudem ermöglicht EUREMA komplexe Lösungskonzepte, die mehrere Feedbackschle

Robustness is the ability to survive unforeseen circumstances without undue damage or loss of function. It has become a requirement expressed in modern building codes, mostly without much advice as to how it can be achieved. Engineering has developed some approaches based on traditional practice as well as recent insight. However, knowledge about robustness remains scattered and ambiguous, making it difficult to apply to many specific cases. The author's attempt to collect and review elements, methods and strategies toward structural robustness, using a holistic, almost philosophical approach. This leads to a set of considerations to guide selection and implementation of measures in specific cases, followed by a collection of applications and examples from the authors practice. The world, engineering and construction are imperfect and not entirely predictable. Robustness provides a measure of structural safety beyond traditional codified design rules.

In order to support a sustainable use of water resources in hard rock aquifers, appropriate experimental and modeling techniques are essential. Multi continuum modeling as a compromise between detailed discrete modeling and rough estimations by analytical solutions, offers a solution by neglecting discrete effects while considering different hydraulic properties. In the scope of this work, characterization techniques are investigated in order to support the choice of an appropriate multi continuum model. Furthermore, sensitivity analyses for multi continuum models are developed to identify the relevant parameters and to explain the dominant flow and transport processes. In order to assess the transferability of insights gained from gas tracer experiments as performed within the Aquifer Analogue Project to water-saturated conditions, both numerical investigations for the water- and the gas-saturated case are analyzed. A double continuum model of the sandstone block investigated in the scope of the Aquifer Analogue Project is developed. Due to strong heterogeneities of the sandstone block, the flow and transport behavior cannot be represented satisfactorily. The model performance could be enhanced by adding discrete elements in order to account for the strong discontinuities found at the field site. A prognosis of the behavior of the water-saturated sandstone block would then be possible. This would create a link between the numerical analyses in this thesis to a real aquifer environment. The applicability of the approach to identify an appropriate model type for a system by means of key-figures to the two-dimensional case is investigated. It is demonstrated that due to the special characteristics of the dipole flow field, the tracer breakthrough curves for different model types are similar. It is thus more difficult than in the one-dimensional case to deduce the number and type of the components contributing to the transport processes from the shape of the tracer breakthrough curve. By means of the sensitivity analyses carried out in the scope of this thesis, the influence of different components of a porous medium is quantified and conclusions concerning the characteristic key-figures are drawn.