论文信息 - A Multi Level Monte Carlo Algorithm for the Treatment of Geometrical and Operational Uncertainties in Internal and External Aerodynamics

A Multi Level Monte Carlo Algorithm for the Treatment of Geometrical and Operational Uncertainties in Internal and External Aerodynamics

In this work we present an efficient Multi Level Monte Carlo (MLMC) algorithm capable of propagating operational uncertainties, due to atmospheric variability of external aerodynamics flows and geometrical uncertainties due to manufacturing tolerances and fatigue. The technique is tested on the NACA 0012 airfoil in transonic conditions and a two compression ramps hypersonic scramjet intake configuration. The efficiency of the proposed MLMC algorithm in terms of accuracy versus overall computational cost is compared to a standard MC method.

Pénélope Leyland | Fabio Nobile | Michele Pisaroni

[1] O. L. Maître,et al. Uncertainty propagation in CFD using polynomial chaos decomposition , 2006 .

[2] Stefan Heinrich,et al. Monte Carlo Complexity of Parametric Integration , 1999, J. Complex..

[3] Michael B. Giles,et al. Multilevel Monte Carlo Path Simulation , 2008, Oper. Res..

[4] Elisabeth Ullmann,et al. Further analysis of multilevel Monte Carlo methods for elliptic PDEs with random coefficients , 2012, Numerische Mathematik.