MODELUNCERTAINTY.PDFVIP

MODEL UNCERTAINTY Marc P. Mignolet Department of Mechanical and Aerospace Engineering Arizona State University Tempe, AZ 85287-6106 Ira A. Fulton School of Engineering Department of Mechanical and Aerospace Engineering Structural Dynamics Group PLAN * Uncertainty: parameter uncertainty - model uncertainty * Benefits of Reduced Order Models * Entropy Maximization * Model Identification within an Uncertain Framework Ira A. Fulton School of Engineering Department of Mechanical and Aerospace Engineering Structural Dynamics Group PARAMETER UNCERTAINTY - MODEL UNCERTAINTY = two sources of uncertainty preventing a perfect match of clean experimental measurements (no measurements uncertainty) with computational predictions: Uncertainty on the parameters of the computational Model: * variations (from part to part) of the values entered in the computational model, e.g. Young’s modulus, Poisson’s ratio, density, dimensions, etc. * Test article and computational model are different parts from the same stock Ira A. Fulton School of Engineering Department of Mechanical and Aerospace Engineering Structural Dynamics Group PARAMETER UNCERTAINTY - MODEL UNCERTAINTY Uncertainty (lack of “realism” ) of the computational model: * computational geometry approximates certain features of the physical model, e.g. _ fasteners (rivets, welds, bolts, lap joints,…), _ plate/beam models of slender components, … _ no warping, out of straight,… * boundary conditions typically differ from those in test/structure * constitutive behavior is modeled: use of linear structural damping isotropic or orthotropic properties, linear elasticity,… Ira A. Fulton School of Engi