Management of Uncertainties

This WP will generate the theoretical backbone of the NODESIM-CFD project, as it will concentrate on the development of the most relevant non-deterministic methodologies. The “research” partners will develop and provide tested and validated software modules for the implemented methodology. The main objectives of this WP3 is the development of innovative mathematical and algorithmic methodologies for a probabilistic approach of the CFD simulations due to the stochastic nature of physical and geometrical parameters, initial and boundary conditions, modelling assumptions and constants.

The following methods are to be developed, assessed and validated for subsequent application to practical cases of industrial relevance:

• Task 3.1: Perturbation techniques whereby the uncertainty variables of the problem are analysed for their sensitivity level, based on “adjoint” methods and development of sensitivity-based methodologies

• Task 3.2: Monte Carlo methods for complete statistics, including surrogate models (polynomial, Kriging, neural network), and DOE and Gaussian processes methods for significant CPU cost reduction.

Overview | Work Package 1 | Work Package 2 | Work Package 3 | Work Package 4 | Work Package 5