EDF - Background

The Sicomore project tends to improve the quality and the reliability of a dynamics simulation using possibly measurement results, both for design or expertise purpose.

The four associated objectives are the following :

• Develop confident tools relative to simulation-measurement correlation, data expansion, structural modification.
• Evaluate the dynamics simulation robustness: error propagation on an inverse problem, error indicators in dynamics, measure of the a posteriori variability of a simulated result (SICODYN benchmark).
• Develop new tools relative to data assimilation and updating.

That can be declined within the following items :

• for design purpose:
• build a confident model of the structure alone;
• model the coupling with its environment;
• model the damping;
• know or identify the excitations;
• have an idea of the error sources, model the uncertainties.
• for expertise purpose:
• build generic models taking into account realisation variabilities ;
• predict the effects of a structural modification; and- use measurement results to improve the model.

The objective of the benchmark is to quantify the variability of dynamic analysis on an industrial structure. The means are blind simulations (pure numerical) and updating simulations taking into account parameters issued from measurements obtained on a real structure.

The study results in statistical estimators which describe:

• The variability due to model parameters.
• The variability due to modelling, computation software, the operator.
• The variability due to measurements (several acquisition systems, several identification methods).
• The variability due to realisation.

These estimators will be re-used on other calculations of the same type for modelling of uncertainties.