Efficient and robust uncertainty quantification for computational fluid dynamics and fluid-structure interaction

Physical uncertainties due to atmospheric variations and production tolerances can nowadays have a larger effect on the accuracy of computational predictions than numerical errors. It is essential to quantify the effect of these uncertainties for reducing design safety factors and robust design optimization. This eventually contributes to the development of aerodynamically more efficient and environmentally friendly transportation and renewable energy technologies.In this thesis, efficient and robust uncertainty quantification methods are developed for computationally intensive flow and fluid-structure interaction simulations including discontinuities and unsteadiness. The proposed methods satisfy the total variation diminishing and extrema diminishing concepts extended to probability space and result in practice in a constant error in time. The considered applications demonstrate that the developed methods have the potential to advance uncertainty quantification for computationally intensive unsteady problems with discontinuities from practically impossible to a routine analysis. The examples also illustrate that taking physical uncertainties into account is a more reliable design practice than using safety margins in combination with deterministic simulation results

• View at Worldcat

Full Text ( Final Version , 11mb )