Die Triebwerksauslegung verlangt eine genaue Analyse der Unsicherheitsquellen und deren Effekt auf die Lebensdauer und Integrität der Bauteile. Bisher wurden diese über die Verwendung von Sicherheitsfaktoren abgedeckt. Mittlerweile ist es aufgrund von HPC (high performance computing) möglich, probabilistische Methoden heranzuziehen, um die Auswirkung von Unsicherheiten der Eingangsgrößen auf die Ausgangsgrößen zu quantifizieren. In dieser Arbeit wird eine speziell auf das Design einer Niederdruckturbine (NDT) zugeschnittene Methode entwickelt, die über eine Kombination von Auslegungs- und Optimierungsmethoden die Unsicherheiten berücksichtigt.      «

Die Triebwerksauslegung verlangt eine genaue Analyse der Unsicherheitsquellen und deren Effekt auf die Lebensdauer und Integrität der Bauteile. Bisher wurden diese über die Verwendung von Sicherheitsfaktoren abgedeckt. Mittlerweile ist es aufgrund von HPC (high performance computing) möglich, probabilistische Methoden heranzuziehen, um die Auswirkung von Unsicherheiten der Eingangsgrößen auf die Ausgangsgrößen zu quantifizieren. In dieser Arbeit wird eine speziell auf das Design einer Niederdruc...     »

Jet engine design requires an accurate analysis of uncertainty sources and of their effect on life and integrity of the parts. Fuel consumptions and noise emissions must be strongly reduced while fulfilling safety requirements. In order to comply with both requisites, the dependence of the system on uncertainty must be diminished. This has been achieved, until now, through the use of safety factors. However, the advent of high power computing now allows the usage of probabilistic methods to determine the effect of input variation on output variables. The aim of this thesis is to develop a new and problem-adapted methodology, based on a combination of appropriate methods for the design and optimization under uncertainty of a low pressure turbine (LPT) rotor. The physical system is described and simulated through finite element methods (FEM). The proposed approach is discussed via the following investigations: possible sources of aleatoric uncertainty in the design process are identified, for instance in engine-to-engine variations, in the manufacturing process and in ambient conditions. The effect of these variations on the responses is measured through sensitivity analysis and uncertainty quantification. The optimization is performed on a reduced design space, which is obtained from the sensitivity analysis. In order to include robustness in this optimization, different architectures are compared and the most efficient, in terms of computational time, is chosen. The use of probabilistic methods provides the practitioner with extensive information on the system, which cannot be gained through the use of classical deterministic techniques. An optimal solution, which satisfies the safety requirements and is insensitive to variation in the design parameters, is obtained by including robustness in the optimization. The methods analyzed here are illustrated for two different cases. In the first, the secondary air system as stand-alone is presented. In the second, a coupled low-thermomechanical model is analyzed for a steady state condition and a reduced transient mission, enabling the calculation of a probabilistic life prediction.     «

Jet engine design requires an accurate analysis of uncertainty sources and of their effect on life and integrity of the parts. Fuel consumptions and noise emissions must be strongly reduced while fulfilling safety requirements. In order to comply with both requisites, the dependence of the system on uncertainty must be diminished. This has been achieved, until now, through the use of safety factors. However, the advent of high power computing now allows the usage of probabilistic methods to dete...     »