Minas Spiridonakos

Research

Handling Uncertainty in Structural System Modeling
Engineering structures are by default operating within a continuously changing environment. Moreover, our knowledge of both their properties and their operational conditions may be defined only in terms of a probabilistic framework. Thus, due to reasons related with their material properties, manufacturing or construction processes, ageing processes, loading conditions, boundary conditions, measurement errors, and others, almost every structural system is characterized by uncertainty. The propagation of uncertainty through the structural system gives rise to corresponding uncertainties of the structural dynamics and in general the behavior of the structure. As a result, we may only have a limited degree of confidence in the condition, reliability and safety of a structure through its whole life cycle. For this reason, the necessity for developing dynamic structural models able to additionally encompass the aforementioned uncertainties has been already highlighted by the international research community over the last years.
The main aim of this study is the treatment of both structural and excitation uncertainties by developing a time-series model with random parameters utilized for the description of uncertainty propagation. The introduced models are able to describe both types of uncertainties by the expansion of their random parameters onto polynomial chaos basis. This approach will rely upon the approximation of the random response of a real world structure or a corresponding analytical model by a suitably defined finite-dimensional polynomial chaos basis. The final objective of the study is twofold: (i) Firstly, the handling of uncertainty in large-scale numerical models that may be used for design optimization of new structures toward a more reliable response even for extreme loading conditions such as earthquakes, or strong wind loads, and (ii) the representation of existing structures by means of a reduced order black-box model that will be able to describe the structural dynamics for a great percentage of its operational spectrum.